[GRASS5] Re: Code freeze

GRASS GRASS developer
1

David (hi all),

On Mon, Oct 16, 2000 at 04:24:28PM +0000, David D Gray wrote:

Markus Neteler wrote:
>
> On Mon, Oct 16, 2000 at 04:07:05PM +0000, David D Gray wrote:

[...]

> What do you think: How much time shall we spend on code freeze? Until
> BUGS is empty? Or definitively release the stable End of October? Or
> earlier?
> Markus

Hi Markus

Perhaps we should do what the Debian folks do: identify certain bugs as
release-critical and then empty that list before release. maybe we take
about a week to assemble such a list, and then publish it encouraging
developers to concentrate on these bugs (a `bug-squash'), and wait till
then to assess the probable time scale.

this sounds very reasonable. I propose that every developer checks
the BUGS file in CVS, update it and add
- "release-critical" if so
- optional add his/her name if wanting to fix it
- remove already fixed bugs (any left?)
Hopefully we can finish updating this list until thursday?
Then I will sort the list into the two sections uncritical and
release-critical and publish it.
Then we go ahead for emptying the list.

Is that o.k. with all developers?

On another point, I will be uploading the v.in.mif later today. It is
working for point/line data (that I have), so there is something for
people to look at, but still needs some work done. I also want to get
v.in.shape tidied up before the release. How will the CVS branch after
the freeze, I mean will there be a frozen branch for bugfix work to go
on, and a devel branch for development to continue?

As a major code-reorganization is required for modularizing the sources
and adding the new auto-conf system I suggest to check-in a completely
new tree. Moving many directories in CVS using CVS commands will be a
pain. Therefore (I hope, Bernhard agrees) we should build a completely
new grass 5.1 source tree.

The branch freeze is our agreement. I think I cannot physically freeze
it to limit submissions to bugfixes. But the new tree shall be established
asap (after discussing it's final form, a draft is here:
documents/new_directory_structure.txt )

Hoping for getting out the stable release soon,

Markus

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

2

Markus Neteler wrote:

[...]
The branch freeze is our agreement. I think I cannot physically freeze
it to limit submissions to bugfixes. But the new tree shall be established
asap (after discussing it's final form, a draft is here:
documents/new_directory_structure.txt )

Hoping for getting out the stable release soon,

Markus

Markus, everyone

It might be a good idea to have at least one pre-release ( GRASS 5.0
PR1) out before the definitive final release, so that users can test the
distribution. Can I suggest a time-scale something like:

Mid-November: release candidate distributed.
(all well) 1 Jan: GRASS 5.0 stable (GRASS ME :slight_smile: )

Of course it depends on release-critical bugs, more might be revealed by
a candidate release.

David

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

3

Hi David, hi all,
On Tue, Oct 17, 2000 at 08:43:39PM +0000, David D Gray wrote:

Markus Neteler wrote:
>
> [...]
> The branch freeze is our agreement. I think I cannot physically freeze
> it to limit submissions to bugfixes. But the new tree shall be established
> asap (after discussing it's final form, a draft is here:
> documents/new_directory_structure.txt )
>
> Hoping for getting out the stable release soon,
>
> Markus

Markus, everyone

It might be a good idea to have at least one pre-release ( GRASS 5.0
PR1) out before the definitive final release, so that users can test the
distribution.

yes, I agree. In fact I was discussing such release candidate with
Bernhard (I should have posted that).

Can I suggest a time-scale something like:

Mid-November: release candidate distributed.
(all well) 1 Jan: GRASS 5.0 stable (GRASS ME :slight_smile: )

Of course it depends on release-critical bugs, more might be revealed by
a candidate release.

Do you think we need so much time? As many new changes are waiting,
I hope we can get out GRASS 5stable much earlier. Or am I wrong?
I thought we could start the new development tree in November...

Markus

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

4

Markus Neteler wrote:

Hi David, hi all,
On Tue, Oct 17, 2000 at 08:43:39PM +0000, David D Gray wrote:
> Markus Neteler wrote:
> >
> > [...]
> > The branch freeze is our agreement. I think I cannot physically freeze
> > it to limit submissions to bugfixes. But the new tree shall be established
> > asap (after discussing it's final form, a draft is here:
> > documents/new_directory_structure.txt )
> >
> > Hoping for getting out the stable release soon,
> >
> > Markus
>
> Markus, everyone
>
> It might be a good idea to have at least one pre-release ( GRASS 5.0
> PR1) out before the definitive final release, so that users can test the
> distribution.
yes, I agree. In fact I was discussing such release candidate with
Bernhard (I should have posted that).

>Can I suggest a time-scale something like:
>
> Mid-November: release candidate distributed.
> (all well) 1 Jan: GRASS 5.0 stable (GRASS ME :slight_smile: )
>
> Of course it depends on release-critical bugs, more might be revealed by
> a candidate release.
Do you think we need so much time? As many new changes are waiting,
I hope we can get out GRASS 5stable much earlier. Or am I wrong?
I thought we could start the new development tree in November...

I can't tell how much time we need, but this is just a guess. I think we
can start the devel tree whenever the freeze is announced, then people
know whether what they are doing goes in the frozen tree or the devel
tree. they are then independent in time. We try to get bug-fixes and
todos fixed as much as possible before a candidate release (but it's not
necessary to fix everything by then). On the feedback, we re-assess the
release-critical stuff at that stage, fix it , then release.

regards

David

Markus

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'